Accurate evaluation of the temporomandibular joint is essential for the differential diagnosis of facial pain patients. A comprehensive evaluation of the temporomandibular joint would ideally provide information regarding the integrity of both hard and soft tissue components. Hard tissue information is obtained from traditional radiographic techniques, while magnetic resonance imaging has the capability to provide the necessary soft tissue information. However, a standardized methodology does not yet exist to optimally acquire, process and interpret magnetic resonance images of the temporomandibular joint. With this in mind, a research program is proposed to develop the foundations of a methodology by optimizing magnetic resonance imaging system parameters and subsequently implementing image processing techniques to further aid in localizing and characterizing certain anatomical structures in the temporomandibular joint. The results of this research will therefore make a valuable contribution towards differential diagnosis of a complex health care problem. This interdisciplinary research, which combines emerging technologies in medical diagnostics, computer engineering and image understanding, will be carried out by a unique collaboration between dental researchers, radiologists, and engineers. The proposed research will enable these investigators to establish the basis on which subsequent extramural funding will be sought.